In [1]:
import yfinance as yf
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns
import datetime
In [2]:
# Stock price prediction uing Monte Carlo simulation
# We will be using Yahoo Finance to obtain historical stock price data

# Stocks included
# Technology stocks
# AAPL - Apple Inc
# MSFT - Microsoft
# GOOGL - Google
# NVDA - NVIDIA 
# TSLA - Tesla
# IBM - IBM Corporation

# Banks:
# JPM - JPMorgan Chase
# BAC - Bank of America 
# WFC - Wells Fargo & Company
# C - Citigroup Inc.
# GS - Goldman Sachs
# WFC - Wells Fargo

# Consumer Goods:
# PG - Procter & Gamble
# KO - Coca-Cola
# PEP - Pepsi
# NKE - Nike
# MCD - McDonald's
# JNJ - Johnson & Johnson

# Energy:
# XOM - Exxon Mobil
# CVX - Chevron
# SHEL - Shell plc
# BP - British Petrol
# TTE - TotalEnergies
# NEE - NextEra Energy

# Retail Trade:
# AMZN - Amazon
# WMT - Walmart
# HD - Home Depot
# TGT - Target 
# COST - Costco
# BABA- Alibaba

pd.options.display.float_format = '{:,.2f}'.format  # Set float format
pd.options.display.max_rows = None
In [3]:
# List of chosen stocks with sectors
stocks = [
    {"ticker": "AAPL", "sector": "Technology"},
    {"ticker": "MSFT", "sector": "Technology"},
    {"ticker": "GOOGL", "sector": "Technology"},
    {"ticker": "NVDA", "sector": "Technology"},
    {"ticker": "TSLA", "sector": "Technology"},
    {"ticker": "IBM", "sector": "Technology"},

    {"ticker": "JPM", "sector": "Banks"},
    {"ticker": "BAC", "sector": "Banks"},
    {"ticker": "WFC", "sector": "Banks"},
    {"ticker": "GS", "sector": "Banks"},
    {"ticker": "MS", "sector": "Banks"},
    {"ticker": "C", "sector": "Banks"},

    {"ticker": "PG", "sector": "Consumer Goods"},
    {"ticker": "KO", "sector": "Consumer Goods"},
    {"ticker": "PEP", "sector": "Consumer Goods"},
    {"ticker": "NKE", "sector": "Consumer Goods"},
    {"ticker": "MCD", "sector": "Consumer Goods"},
    {"ticker": "JNJ", "sector": "Consumer Goods"},

    {"ticker": "XOM", "sector": "Energy"},
    {"ticker": "CVX", "sector": "Energy"},
    {"ticker": "SHEL", "sector": "Energy"},
    {"ticker": "BP", "sector": "Energy"},
    {"ticker": "TTE", "sector": "Energy"},
    {"ticker": "NEE", "sector": "Energy"},

    {"ticker": "AMZN", "sector": "Retail Trade"},
    {"ticker": "WMT", "sector": "Retail Trade"},
    {"ticker": "HD", "sector": "Retail Trade"},
    {"ticker": "TGT", "sector": "Retail Trade"},
    {"ticker": "COST", "sector": "Retail Trade"},
    {"ticker": "BABA", "sector": "Retail Trade"}
]

# Empty DataFrame to store data
combined_data = pd.DataFrame()

for stock in stocks:
    # Fetch historical data from Yahoo Finance
    stock_data = yf.download(stock["ticker"], start='1990-01-01', end='2024-05-22',progress=False)
    
    # Add Ticker and Sector columns 
    stock_data['Ticker'] = stock["ticker"]
    stock_data['Sector'] = stock["sector"]
    
    # Append data to df
    combined_data = pd.concat([combined_data, stock_data])

# Reset index and save data
combined_data.reset_index(inplace=True)
combined_data.to_csv('historical_stock_data.csv', index=False)
In [4]:
# Let's check the status of our dataframe
combined_data.head()
Out[4]:
Date Open High Low Close Adj Close Volume Ticker Sector
0 1990-01-02 0.31 0.33 0.31 0.33 0.26 183198400 AAPL Technology
1 1990-01-03 0.34 0.34 0.33 0.33 0.26 207995200 AAPL Technology
2 1990-01-04 0.34 0.35 0.33 0.34 0.27 221513600 AAPL Technology
3 1990-01-05 0.34 0.34 0.33 0.34 0.27 123312000 AAPL Technology
4 1990-01-08 0.33 0.34 0.33 0.34 0.27 101572800 AAPL Technology
In [5]:
# Let's check our data types
combined_data.dtypes
Out[5]:
Date         datetime64[ns]
Open                float64
High                float64
Low                 float64
Close               float64
Adj Close           float64
Volume                int64
Ticker               object
Sector               object
dtype: object
In [6]:
# Let's have a quick look at our summary statistics
combined_data.describe()

# We have 237k datapoints
# We have a date range of 1990 to 2024
# Average closing price is $62.26 throughout all stock, where closing prices range from 0.07 to 953 USD.
Out[6]:
Date Open High Low Close Adj Close Volume
count 237039 237,039.00 237,039.00 237,039.00 237,039.00 237,039.00 237,039.00
mean 2008-01-18 06:46:21.932087296 62.66 63.31 61.99 62.66 48.26 34,280,793.52
min 1990-01-02 00:00:00 0.07 0.07 0.07 0.07 0.07 0.00
25% 1999-10-27 00:00:00 17.19 17.38 17.00 17.19 10.10 3,669,700.00
50% 2008-06-02 00:00:00 40.53 40.99 40.06 40.53 24.59 7,665,400.00
75% 2016-07-15 00:00:00 72.53 73.29 71.85 72.57 53.53 20,573,050.00
max 2024-05-21 00:00:00 958.51 974.00 941.03 953.86 953.86 7,421,640,800.00
std NaN 78.06 78.90 77.18 78.06 69.04 106,244,844.09
In [7]:
# Check missing values for each column
missing_values_per_column = combined_data.isnull().sum()
print("Number of missing values for each column:")
print(missing_values_per_column)

Number of missing values for each column:
Date         0
Open         0
High         0
Low          0
Close        0
Adj Close    0
Volume       0
Ticker       0
Sector       0
dtype: int64
In [8]:
# Taking averages of each sector for comparison
numeric_columns = ['Open', 'High', 'Low', 'Close', 'Adj Close', 'Volume']
sector_aggregated_data = combined_data.groupby(['Sector'])[numeric_columns].mean().reset_index()

# Save aggregated data to CSV
sector_aggregated_data.to_csv('sector_aggregated_data.csv', index=False)
sector_aggregated_data.head()
Out[8]:
Sector Open High Low Close Adj Close Volume
0 Banks 79.63 80.59 78.61 79.61 58.11 20,718,479.68
1 Consumer Goods 58.39 58.87 57.92 58.41 45.81 8,112,461.54
2 Energy 47.65 48.06 47.23 47.66 28.35 6,556,391.77
3 Retail Trade 68.74 69.46 68.01 68.76 60.47 31,212,821.91
4 Technology 60.19 60.93 59.44 60.21 51.61 122,082,322.48
In [9]:
# Plotting closing prices for each stock, compared to other stocks in the sector
sectors = combined_data['Sector'].unique()

for sector in sectors:
    sector_stocks = [stock for stock in stocks if stock['sector'] == sector]

    plt.figure(figsize=(8, 4))
    plt.title(f'Closing Prices of {sector} Stocks')
    plt.xlabel('Date')
    plt.ylabel('Closing Price (USD)')

    for stock in sector_stocks:
        stock_data = combined_data[combined_data['Ticker'] == stock['ticker']]
        plt.plot(stock_data['Date'], stock_data['Close'], label=stock['ticker'], linewidth=0.6)

    plt.legend()
    plt.show()

# Some observations
# We can see tech stocks were cheap for very long, they are booming after 2020, NVIDIA especially
# We can see Goldman Sachs and CitiGroup were severaly affected by the 2008 mortgage crisis
# Energy stocks all show a similar trend, with a slight decline around 2008, stable around 2008-2020, and increasing afterwards
# We see that retail stocks are also are on an upward trend after 2020, especially COSTCO
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
In [10]:
# Plot comparing the different sector
plt.figure(figsize=(10, 5))
# Grouping data by sector and date, then calculating the average closing price for each sector
sector_avg_data = combined_data.groupby(['Sector', 'Date'])['Close'].mean().reset_index()

# Plotting average closing prices for each sector
for sector in ['Tech', 'Banks', 'Consumer Goods', 'Energy', 'Retail Trade']:
    sector_data = sector_avg_data[sector_avg_data['Sector'] == sector]
    plt.plot(sector_data['Date'], sector_data['Close'], label=sector,linewidth=0.6)

# Labels and title
plt.xlabel('Date')
plt.ylabel('Average Closing Price (USD)')
plt.title('Average Closing Prices of Different Sectors Over Time')
plt.legend()
plt.grid(True)
plt.xticks(rotation=45)  # Rotate x-axis labels for better readability

plt.show()
# This plot also reinforces our observations from the previous stock graphs
No description has been provided for this image
In [11]:
# Steps of Monte Carlo Simulation
# Get Historical Data: Done
# Calculate Daily Returns: We will do this now
# Simulate Future Prices: Generate multiplandom increments or decrements to the current price, based on the historical volatility of the stock.
# Calculate Metrics
# Visualize & evaluate results
In [12]:
# Calculate daily returns for each stock
for stock in stocks:
    ticker = stock["ticker"]
    combined_data.loc[combined_data['Ticker'] == ticker, 'Daily Return'] = combined_data.loc[combined_data['Ticker'] == ticker, 'Close'].pct_change()
    
# Replace NaN values with if needed
combined_data['Daily Return'] = combined_data['Daily Return'].fillna(0)
In [13]:
combined_data.describe()
Out[13]:
Date Open High Low Close Adj Close Volume Daily Return
count 237039 237,039.00 237,039.00 237,039.00 237,039.00 237,039.00 237,039.00 237,039.00
mean 2008-01-18 06:46:21.932087296 62.66 63.31 61.99 62.66 48.26 34,280,793.52 0.00
min 1990-01-02 00:00:00 0.07 0.07 0.07 0.07 0.07 0.00 -0.52
25% 1999-10-27 00:00:00 17.19 17.38 17.00 17.19 10.10 3,669,700.00 -0.01
50% 2008-06-02 00:00:00 40.53 40.99 40.06 40.53 24.59 7,665,400.00 0.00
75% 2016-07-15 00:00:00 72.53 73.29 71.85 72.57 53.53 20,573,050.00 0.01
max 2024-05-21 00:00:00 958.51 974.00 941.03 953.86 953.86 7,421,640,800.00 0.87
std NaN 78.06 78.90 77.18 78.06 69.04 106,244,844.09 0.02
In [14]:
# Checking if daily returns were calculated correctly
for stock in stocks:
    ticker = stock["ticker"]
    stock_daily_returns = combined_data.loc[combined_data['Ticker'] == ticker, ['Date', 'Daily Return']]
    print(f"Daily Returns for {ticker}:")
    print(stock_daily_returns.head(5))

Daily Returns for AAPL:
        Date  Daily Return
0 1990-01-02          0.00
1 1990-01-03          0.01
2 1990-01-04          0.00
3 1990-01-05          0.00
4 1990-01-08          0.01
Daily Returns for MSFT:
           Date  Daily Return
8663 1990-01-02          0.00
8664 1990-01-03          0.01
8665 1990-01-04          0.03
8666 1990-01-05         -0.02
8667 1990-01-08          0.02
Daily Returns for GOOGL:
            Date  Daily Return
17326 2004-08-19          0.00
17327 2004-08-20          0.08
17328 2004-08-23          0.01
17329 2004-08-24         -0.04
17330 2004-08-25          0.01
Daily Returns for NVDA:
            Date  Daily Return
22299 1999-01-22          0.00
22300 1999-01-25          0.10
22301 1999-01-26         -0.08
22302 1999-01-27         -0.00
22303 1999-01-28         -0.00
Daily Returns for TSLA:
            Date  Daily Return
28673 2010-06-29          0.00
28674 2010-06-30         -0.00
28675 2010-07-01         -0.08
28676 2010-07-02         -0.13
28677 2010-07-06         -0.16
Daily Returns for IBM:
            Date  Daily Return
32171 1990-01-02          0.00
32172 1990-01-03          0.01
32173 1990-01-04          0.01
32174 1990-01-05         -0.00
32175 1990-01-08          0.01
Daily Returns for JPM:
            Date  Daily Return
40834 1990-01-02          0.00
40835 1990-01-03          0.03
40836 1990-01-04          0.00
40837 1990-01-05          0.00
40838 1990-01-08          0.00
Daily Returns for BAC:
            Date  Daily Return
49497 1990-01-02          0.00
49498 1990-01-03          0.01
49499 1990-01-04         -0.02
49500 1990-01-05         -0.02
49501 1990-01-08          0.01
Daily Returns for WFC:
            Date  Daily Return
58160 1990-01-02          0.00
58161 1990-01-03         -0.04
58162 1990-01-04         -0.01
58163 1990-01-05         -0.03
58164 1990-01-08          0.03
Daily Returns for GS:
            Date  Daily Return
66823 1999-05-04          0.00
66824 1999-05-05         -0.02
66825 1999-05-06         -0.02
66826 1999-05-07          0.09
66827 1999-05-10         -0.05
Daily Returns for MS:
            Date  Daily Return
73127 1993-02-23          0.00
73128 1993-02-24          0.03
73129 1993-02-25          0.02
73130 1993-02-26          0.02
73131 1993-03-01         -0.03
Daily Returns for C:
            Date  Daily Return
80995 1990-01-02          0.00
80996 1990-01-03          0.01
80997 1990-01-04         -0.01
80998 1990-01-05          0.01
80999 1990-01-08          0.01
Daily Returns for PG:
            Date  Daily Return
89658 1990-01-02          0.00
89659 1990-01-03         -0.00
89660 1990-01-04         -0.02
89661 1990-01-05         -0.02
89662 1990-01-08          0.02
Daily Returns for KO:
            Date  Daily Return
98321 1990-01-02          0.00
98322 1990-01-03         -0.01
98323 1990-01-04         -0.00
98324 1990-01-05         -0.01
98325 1990-01-08          0.02
Daily Returns for PEP:
             Date  Daily Return
106984 1990-01-02          0.00
106985 1990-01-03         -0.01
106986 1990-01-04         -0.01
106987 1990-01-05         -0.01
106988 1990-01-08          0.02
Daily Returns for NKE:
             Date  Daily Return
115647 1990-01-02          0.00
115648 1990-01-03         -0.02
115649 1990-01-04         -0.02
115650 1990-01-05          0.03
115651 1990-01-08         -0.00
Daily Returns for MCD:
             Date  Daily Return
124310 1990-01-02          0.00
124311 1990-01-03         -0.01
124312 1990-01-04         -0.02
124313 1990-01-05         -0.02
124314 1990-01-08          0.02
Daily Returns for JNJ:
             Date  Daily Return
132973 1990-01-02          0.00
132974 1990-01-03          0.00
132975 1990-01-04          0.00
132976 1990-01-05         -0.01
132977 1990-01-08          0.02
Daily Returns for XOM:
             Date  Daily Return
141636 1990-01-02          0.00
141637 1990-01-03         -0.01
141638 1990-01-04         -0.01
141639 1990-01-05         -0.01
141640 1990-01-08          0.02
Daily Returns for CVX:
             Date  Daily Return
150299 1990-01-02          0.00
150300 1990-01-03         -0.02
150301 1990-01-04         -0.01
150302 1990-01-05         -0.01
150303 1990-01-08          0.01
Daily Returns for SHEL:
             Date  Daily Return
158962 1990-01-02          0.00
158963 1990-01-03         -0.03
158964 1990-01-04          0.00
158965 1990-01-05         -0.02
158966 1990-01-08          0.15
Daily Returns for BP:
             Date  Daily Return
167625 1990-01-02          0.00
167626 1990-01-03          0.00
167627 1990-01-04          0.00
167628 1990-01-05          0.01
167629 1990-01-08          0.01
Daily Returns for TTE:
             Date  Daily Return
176288 1991-10-25          0.00
176289 1991-10-28         -0.02
176290 1991-10-29         -0.01
176291 1991-10-30          0.02
176292 1991-10-31          0.05
Daily Returns for NEE:
             Date  Daily Return
184491 1990-01-02          0.00
184492 1990-01-03         -0.01
184493 1990-01-04         -0.02
184494 1990-01-05         -0.01
184495 1990-01-08         -0.02
Daily Returns for AMZN:
             Date  Daily Return
193154 1997-05-15          0.00
193155 1997-05-16         -0.12
193156 1997-05-19         -0.01
193157 1997-05-20         -0.04
193158 1997-05-21         -0.13
Daily Returns for WMT:
             Date  Daily Return
199953 1990-01-02          0.00
199954 1990-01-03          0.00
199955 1990-01-04         -0.01
199956 1990-01-05         -0.01
199957 1990-01-08          0.01
Daily Returns for HD:
             Date  Daily Return
208616 1990-01-02          0.00
208617 1990-01-03          0.00
208618 1990-01-04          0.01
208619 1990-01-05         -0.01
208620 1990-01-08         -0.01
Daily Returns for TGT:
             Date  Daily Return
217279 1990-01-02          0.00
217280 1990-01-03          0.01
217281 1990-01-04          0.00
217282 1990-01-05          0.00
217283 1990-01-08          0.01
Daily Returns for COST:
             Date  Daily Return
225942 1990-01-02          0.00
225943 1990-01-03          0.02
225944 1990-01-04          0.01
225945 1990-01-05          0.01
225946 1990-01-08         -0.02
Daily Returns for BABA:
             Date  Daily Return
234605 2014-09-19          0.00
234606 2014-09-22         -0.04
234607 2014-09-23         -0.03
234608 2014-09-24          0.04
234609 2014-09-25         -0.02
In [15]:
# Let's also look at min,max,mean daily returns for stocks to ensure validity
returns_statistics = []

for stock in stocks:
    ticker = stock["ticker"]
    sector = stock["sector"]
    stock_daily_returns = combined_data.loc[combined_data['Ticker'] == ticker, 'Daily Return']
    
    max_return = stock_daily_returns.max()
    min_return = stock_daily_returns.min()
    mean_return = stock_daily_returns.mean()
    
    if sector == "Banks":
        sector_label = "BP"
    else:
        sector_label = sector
    
    returns_statistics.append({
        'Stock': ticker,
        'Sector': sector_label,
        'Max Return': max_return,
        'Min Return': min_return,
        'Mean Return': mean_return
    })
returns_statistics_df = pd.DataFrame(returns_statistics)
returns_statistics_df.head()
# Everything seems ok, we can move forward with the next step, which is to simulate future prices
Out[15]:
Stock Sector Max Return Min Return Mean Return
0 AAPL Technology 0.33 -0.52 0.00
1 MSFT Technology 0.20 -0.16 0.00
2 GOOGL Technology 0.20 -0.12 0.00
3 NVDA Technology 0.42 -0.35 0.00
4 TSLA Technology 0.24 -0.21 0.00
In [16]:
# Parameters
num_simulations = 20  # Number of simulations
num_days = 504  # Number of future days to simulate

# Dictionary to store simulated price paths for each stock
simulated_price_paths_per_stock = {}

# Monte Carlo simulation for each stock
for stock in stocks:
    ticker = stock["ticker"]
    
    # Historical daily returns for the current stock
    historical_returns = combined_data.loc[combined_data['Ticker'] == ticker, 'Daily Return'].dropna().values
    
    # Mean and standard deviation of daily returns for the current stock
    mean_return = historical_returns.mean()
    std_return = historical_returns.std()
    
    # Set initial price to the last known price in the historical data for the current stock
    initial_price = combined_data.loc[combined_data['Ticker'] == ticker, 'Close'].iloc[-1]
    
    # Initialize array to store simulated price paths for the current stock
    simulated_price_paths = np.zeros((num_simulations, num_days + 1))  # Increase num_days by 1 to account for initial price
    
    # Perform Monte Carlo simulation for the current stock
    for i in range(num_simulations):
        # Initialize price path for simulation i
        simulated_price_path = np.zeros(num_days + 1)  # Increase num_days by 1 to account for initial price
        simulated_price_path[0] = initial_price
        
        # Generate random daily returns based on historical mean and standard deviation
        random_returns = np.random.normal(mean_return, std_return, num_days)
        
        # Calculate future prices
        for j in range(1, num_days + 1):
            simulated_price_path[j] = simulated_price_path[j - 1] * (1 + random_returns[j - 1])  # Adjust index to start from 0
        
        # Store simulated price path
        simulated_price_paths[i] = simulated_price_path
    
    # Store simulated price paths for the current stock
    simulated_price_paths_per_stock[ticker] = simulated_price_paths

# Plot simulated prices for each stock on separate plots
for ticker, simulated_price_paths in simulated_price_paths_per_stock.items():
    plt.figure(figsize=(8, 3))
    plt.title(f'Simulated Price Paths for {ticker}')
    plt.xlabel('Days')
    plt.ylabel('Price')

    # Generate future dates starting from the last known date, limited to num_days
    last_known_date = combined_data[combined_data['Ticker'] == ticker]['Date'].max()
    future_dates = [last_known_date + datetime.timedelta(days=i) for i in range(1, num_days + 1)]

    # Plot each simulation path
    for i in range(num_simulations):
        plt.plot(future_dates, simulated_price_paths[i][1:], linewidth=0.5, alpha=0.5)  # Exclude the initial price when plotting
    plt.show()
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
In [17]:
# Plot historical and predicted price paths for each stock
for ticker, simulated_price_paths in simulated_price_paths_per_stock.items():
    plt.figure(figsize=(8, 3))

    # Plot historical prices
    historical_data = combined_data[combined_data['Ticker'] == ticker]
    plt.plot(historical_data['Date'], historical_data['Close'], label=f'{ticker} (Historical)', linewidth=1)

    # Aggregate simulated prices for future dates
    last_known_date = historical_data['Date'].values[-1]
    avg_simulated_prices = np.mean(simulated_price_paths[:, 1:], axis=0)  # Exclude the first known price
    future_dates = pd.date_range(start=last_known_date, periods=len(avg_simulated_prices), freq='B')  # Start from the last known date
    plt.plot(future_dates, avg_simulated_prices, linewidth=1, color='red', label=f'{ticker} (Predicted)')

    plt.xlabel('Date')
    plt.ylabel('Price')
    plt.title(f'Historical and Predicted Price Paths for {ticker}')
    plt.legend()
    plt.show()
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
In [18]:
sector_avg_color = 'green'  
unique_sectors = set(stock["sector"] for stock in stocks)

# Predicted prices for each sector
for sector in unique_sectors:
    plt.figure(figsize=(8, 4))
    plt.title(f'Predicted Price Paths for Stocks in {sector}')
    plt.xlabel('Date')
    plt.ylabel('Price')
    
    # Filter stocks belonging to the current sector
    sector_stocks = [stock["ticker"] for stock in stocks if stock["sector"] == sector]

    # Plot sector average
    sector_avg_prices = []
    for ticker in sector_stocks:
        simulated_price_paths = simulated_price_paths_per_stock[ticker]
        last_known_date = combined_data[combined_data['Ticker'] == ticker]['Date'].max()
        avg_simulated_prices = np.mean(simulated_price_paths[:, 1:], axis=0)  # Exclude the first known price    
        # Generate future dates
        future_dates = [last_known_date + datetime.timedelta(days=i) for i in range(1, len(avg_simulated_prices) + 1)]  
        # Plot sector average only for the first stock
        if ticker == sector_stocks[0]:
            plt.plot(future_dates, avg_simulated_prices, label=f'{sector} Average', color=sector_avg_color, linestyle='--')      
        else:
            plt.plot(future_dates, avg_simulated_prices, linestyle='--', color=sector_avg_color)  # Plot without label
        # Store sector average prices for later use
        sector_avg_prices.append(avg_simulated_prices)
    
    # Calculate and plot sector average
    sector_avg_prices = np.mean(sector_avg_prices, axis=0)
    future_dates = [last_known_date + datetime.timedelta(days=i) for i in range(1, len(sector_avg_prices) + 1)]
    plt.plot(future_dates, sector_avg_prices, label=f'{sector} Average', color=sector_avg_color, linestyle='--')

    # Plot predicted prices for each stock in the sector
    for ticker in sector_stocks:
        simulated_price_paths = simulated_price_paths_per_stock[ticker]
        last_known_date = combined_data[combined_data['Ticker'] == ticker]['Date'].max()
        avg_simulated_prices = np.mean(simulated_price_paths[:, 1:], axis=0)  # Exclude the first known price
        
        # Generate future dates
        future_dates = [last_known_date + datetime.timedelta(days=i) for i in range(1, len(avg_simulated_prices) + 1)]
        
        # Plot individual stock's predicted prices
        plt.plot(future_dates, avg_simulated_prices, label=ticker)

    # Move the legend to the right side
    plt.legend(loc='center left', bbox_to_anchor=(1, 0.5))
    
    plt.grid(True)
    plt.xticks(rotation=45)
    plt.tight_layout()
    plt.show()
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
No description has been provided for this image
In [ ]: